CN107452001A - A kind of remote sensing images sequences segmentation method based on improved FCM algorithm - Google Patents

A kind of remote sensing images sequences segmentation method based on improved FCM algorithm Download PDF

Info

Publication number
CN107452001A
CN107452001A CN201710448043.XA CN201710448043A CN107452001A CN 107452001 A CN107452001 A CN 107452001A CN 201710448043 A CN201710448043 A CN 201710448043A CN 107452001 A CN107452001 A CN 107452001A
Authority
CN
China
Prior art keywords
mrow
msub
munderover
cluster
remote sensing
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201710448043.XA
Other languages
Chinese (zh)
Inventor
杜根远
卢涵宇
姚丹丹
邱颖豫
袁雅婧
胡涛
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xuchang University
Original Assignee
Xuchang University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xuchang University filed Critical Xuchang University
Priority to CN201710448043.XA priority Critical patent/CN107452001A/en
Publication of CN107452001A publication Critical patent/CN107452001A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/90—Determination of colour characteristics
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10032—Satellite or aerial image; Remote sensing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30181—Earth observation

Abstract

The invention belongs to remote sensing images technical field, discloses a kind of remote sensing images sequences segmentation method based on improved FCM algorithm, including：Coloured image is transformed into Lab space from RGB color；Secondly the ab components of image pixel are extracted；Then the histogram of ab components is calculated, and the cluster centre and cluster numbers of image are obtained with histogram interaction；FCM clusters finally are carried out using object function, record cluster result.Present procedure run time is shorter, effectively improves the precision and efficiency of image segmentation；Effect and final segmentation effect class center and cluster numbers of the improved FCM partitioning algorithms under different clusters, overcome the randomness that traditional FCM algorithms obtain initial cluster center and cluster numbers, fork entropy distance measure is selected again simultaneously, so that this method is not necessarily dependent on spherical data distribution, cluster segmentation effect is preferable, edge clear, efficiency also greatly improve.

Description

A kind of remote sensing images sequences segmentation method based on improved FCM algorithm
Technical field
The invention belongs to remote sensing images technical field, more particularly to a kind of remote sensing images sequence based on improved FCM algorithm Dividing method.
Background technology
Remote sensing images have been widely used in land use ＆ environment as the significant data source for making and updating the data storehouse With the continuous development of remote sensing and information technique, various magnanimity in the fields such as monitoring, resource, exploration, Disaster Assessment, urban planning Remotely-sensed data can be obtained by earth observation technology, but data are changed among the process of information and still suffered from very More bottlenecks, wherein Remote Sensing Image Segmentation are exactly a very crucial technology, while are also the difficult point and again of image processing field Point, Remote Sensing Image Segmentation refer to handle remote sensing images, analysis, therefrom extract the technology and process of target, although current There is substantial amounts of image segmentation algorithm, but gray level is more, and information content is larger, obscure boundary because remote sensing images generally have It is clear, the features such as target type is more, cause these algorithms many problems to be in actual applications also present, such as ＆ poor for applicability is split Efficiency is low, and segmentation precision is not high.In addition the segmentation of remote sensing images has uncertainty, different application purposes and user in itself Different to the part of interesting image, the level of information for it is expected to obtain from image also tends to difference, causes to be difficult to set up completely Accurate dividing method is split to remote sensing images, all in all, when carrying out Remote Sensing Image Segmentation processing, generally use mould Theoretical method is pasted, the operational effect and initial cluster center and cluster numbers of traditional fuzzy C- averages (FCM) clustering algorithm are close Correlation, and initial cluster center and cluster numbers have randomness, cause each run difference on effect larger；Rely heavily on In spherical sample data, for non-spherical sample data, its Clustering Effect is unsatisfactory.
In summary, the problem of prior art is present be：The each run effect of traditional fuzzy C- averages (FCM) clustering algorithm Fruit differs greatly；Spherical sample data is largely dependent upon, its Clustering Effect is unsatisfactory for non-spherical sample data.
The content of the invention
The problem of existing for prior art, the invention provides a kind of remote sensing images sequence based on improved FCM algorithm Dividing method.
The present invention is achieved in that a kind of remote sensing images sequences segmentation method based on improved FCM algorithm, described to be based on The remote sensing images sequences segmentation method of improved FCM algorithm comprises the following steps：
Step 1, coloured image is transformed into Lab space from RGB color；
Step 2, secondly extract the ab components of image pixel；
The histogram of step 3, then calculating ab components, and the cluster centre of image is obtained with gathering with histogram interaction Class number；
Step 4, FCM clusters finally are carried out using object function, record cluster result.
Further, the object function is：
Wherein, α is to represent supervision and do not supervise the parameter of degree, it can with the ratio of non-label and the sample of label come Represent, a balance is maintained between supervising and not supervising；bjIt is Boolean type variable, is marked with it label and non-label Sample；If bjFor 0, then it represents that sample XjIt is non-label；If bjFor 1, then it represents that sample XjIt is label；Therefore The degree of membership of exemplar represents F=[f with a matrix formij]；Distance in formula still takes Mahalanobis above Distance；M=2 is taken, iterative formula is：
Advantages of the present invention and good effect are：During cluster, by carrying out similitude with the sample of label Compare, the degree of accuracy of algorithm cluster can be improved；Remote Sensing Image Segmentation effect after splitting with improved FCM clustering algorithms is bright Aobvious, program runtime is shorter, the precision and efficiency of image segmentation can be effectively improved, because improved FCM algorithms Initial poly- figure, effect and final segmentation effect of the improved FCM partitioning algorithms under different clusters are obtained using image histogram Fruit center and cluster numbers, overcome traditional FCM algorithms and obtain the randomness of initial cluster center and cluster numbers, while select again Fork entropy distance measure so that this method is not necessarily dependent on spherical data distribution, cluster segmentation effect is preferable, edge clear, effect Rate also greatly improves.
Brief description of the drawings
Fig. 1 is the remote sensing images sequences segmentation method flow diagram provided in an embodiment of the present invention based on improved FCM algorithm.
Embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, with reference to embodiments, to the present invention It is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not used to Limit the present invention.
The application principle of the present invention is explained in detail below in conjunction with the accompanying drawings.
As shown in figure 1, the remote sensing images sequences segmentation method bag provided in an embodiment of the present invention based on improved FCM algorithm Include：
S101：Coloured image is transformed into Lab space from RGB color；
S102：Secondly the ab components (chrominance space) of image pixel are extracted；
S103：Then the histogram of ab components is calculated, and the cluster centre and cluster of image are obtained with histogram interaction Number；
S104：FCM clusters finally are carried out using object function, record cluster result.
Object function is in step S104：
Wherein, α is to represent supervision and do not supervise the parameter of degree, it can with the ratio of non-label and the sample of label come Represent, a balance is maintained between supervising and not supervising.bjIt is Boolean type variable, is marked with it label and non-label Sample.If bjFor 0, then it represents that sample XjIt is non-label；If bjFor 1, then it represents that sample XjIt is label；Therefore The degree of membership of exemplar represents F=[f with a matrix formij].Distance in formula still takes Mahalanobis above Distance.M=2 is taken, iterative formula is：
It is explained in detail with reference to the application effect to comparing the present invention.
The traditional FCM clustering algorithms of table 1 and the contrast of the present invention
Remote Sensing Image Segmentation positive effect after splitting with improved FCM clustering algorithms, program runtime is shorter, energy The precision and efficiency of image segmentation are enough effectively improved, because improved FCM algorithms are obtained initially using image histogram Poly- figure, effect and final segmentation effect class center and cluster numbers of the improved FCM partitioning algorithms under different clusters, overcomes Traditional FCM algorithms obtain the randomness of initial cluster center and cluster numbers, while have selected fork entropy distance measure again so that the party Method is not necessarily dependent on spherical data distribution, and cluster segmentation effect is preferable, and edge clear, efficiency also greatly improves.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all essences in the present invention All any modification, equivalent and improvement made within refreshing and principle etc., should be included in the scope of the protection.

Claims (2)

  1. A kind of 1. remote sensing images sequences segmentation method based on improved FCM algorithm, it is characterised in that described to be calculated based on improvement FCM The remote sensing images sequences segmentation method of method comprises the following steps：
    Step 1, coloured image is transformed into Lab space from RGB color；
    Step 2, secondly extract the ab components of image pixel；
    Step 3, the histogram of ab components is then calculated, and the cluster centre and cluster numbers of image are obtained with histogram interaction；
    Step 4, FCM clusters finally are carried out using object function, record cluster result.
  2. 2. the remote sensing images sequences segmentation method based on improved FCM algorithm as claimed in claim 1, it is characterised in that described Object function is：
    <mrow> <msub> <mi>J</mi> <mi>m</mi> </msub> <mrow> <mo>(</mo> <mi>U</mi> <mo>,</mo> <mi>V</mi> <mo>)</mo> </mrow> <mo>=</mo> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>j</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </munderover> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>c</mi> </munderover> <msup> <msub> <mi>u</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <mi>m</mi> </msup> <msup> <msub> <mi>d</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <mn>2</mn> </msup> <mo>+</mo> <mi>&amp;alpha;</mi> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>j</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </munderover> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>c</mi> </munderover> <msup> <mrow> <mo>(</mo> <msub> <mi>u</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <mo>-</mo> <msub> <mi>f</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <msub> <mi>b</mi> <mi>j</mi> </msub> <mo>)</mo> </mrow> <mi>m</mi> </msup> <msup> <msub> <mi>d</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <mn>2</mn> </msup> <mo>;</mo> </mrow>
    Wherein, α is the parameter for representing supervision and not supervising degree, and it can be with the ratio of non-label and the sample of label come table Show, a balance is maintained between supervising and not supervising；bjIt is Boolean type variable, with its sample with non-label to mark label This；If bjFor 0, then it represents that sample XjIt is non-label；If bjFor 1, then it represents that sample XjIt is label；Therefore marked The degree of membership of signed-off sample sheet represents F=[f with a matrix formij]；Distance in formula still take Mahalanobis above away from From；M=2 is taken, iterative formula is：
    <mrow> <msub> <mi>U</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <mo>=</mo> <mfrac> <mn>1</mn> <mrow> <mn>1</mn> <mo>+</mo> <mi>&amp;alpha;</mi> </mrow> </mfrac> <mrow> <mo>&amp;lsqb;</mo> <mrow> <mfrac> <mrow> <mn>1</mn> <mo>+</mo> <mrow> <mo>(</mo> <mrow> <mn>1</mn> <mo>-</mo> <msub> <mi>b</mi> <mi>j</mi> </msub> <munderover> <mi>&amp;Sigma;</mi> <mrow> <mi>k</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>c</mi> </munderover> <msub> <mi>f</mi> <mrow> <mi>K</mi> <mi>j</mi> </mrow> </msub> </mrow> <mo>)</mo> </mrow> </mrow> <mrow> <munderover> <mi>&amp;Sigma;</mi> <mrow> <mi>k</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>c</mi> </munderover> <mfrac> <msubsup> <mi>d</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mn>2</mn> </msubsup> <msubsup> <mi>d</mi> <mrow> <mi>k</mi> <mi>j</mi> </mrow> <mn>2</mn> </msubsup> </mfrac> </mrow> </mfrac> <mo>+</mo> <msub> <mi>&amp;alpha;f</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <msub> <mi>b</mi> <mi>j</mi> </msub> </mrow> <mo>&amp;rsqb;</mo> </mrow> <mo>;</mo> </mrow>
    <mrow> <msub> <mi>V</mi> <mi>i</mi> </msub> <mo>=</mo> <mfrac> <mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>j</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </munderover> <msubsup> <mi>u</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mn>2</mn> </msubsup> <msub> <mi>X</mi> <mi>j</mi> </msub> </mrow> <mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>j</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </munderover> <msubsup> <mi>u</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mn>2</mn> </msubsup> </mrow> </mfrac> <mo>.</mo> </mrow> 1
CN201710448043.XA 2017-06-14 2017-06-14 A kind of remote sensing images sequences segmentation method based on improved FCM algorithm Pending CN107452001A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710448043.XA CN107452001A (en) 2017-06-14 2017-06-14 A kind of remote sensing images sequences segmentation method based on improved FCM algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710448043.XA CN107452001A (en) 2017-06-14 2017-06-14 A kind of remote sensing images sequences segmentation method based on improved FCM algorithm

Publications (1)

Publication Number Publication Date
CN107452001A true CN107452001A (en) 2017-12-08

Family

ID=60486789

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710448043.XA Pending CN107452001A (en) 2017-06-14 2017-06-14 A kind of remote sensing images sequences segmentation method based on improved FCM algorithm

Country Status (1)

Country Link
CN (1) CN107452001A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108830297A (en) * 2018-05-19 2018-11-16 烟台大学 A kind of multi-spectrum remote sensing image terrain classification method
CN111681245A (en) * 2020-06-17 2020-09-18 中原工学院 Method for segmenting remote sensing image based on tree structure adaptive weight k-means algorithm
CN111754501A (en) * 2020-06-30 2020-10-09 重庆师范大学 Self-adaptive soil image shadow detection method based on FCM algorithm

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102254303A (en) * 2011-06-13 2011-11-23 河海大学 Methods for segmenting and searching remote sensing image
CN102880872A (en) * 2012-08-28 2013-01-16 中国科学院东北地理与农业生态研究所 Classification and construction method for semi-supervised support vector machine (SVM) remote sensing image
CN104134219A (en) * 2014-08-12 2014-11-05 吉林大学 Color image segmentation algorithm based on histograms
CN105512622A (en) * 2015-12-01 2016-04-20 北京航空航天大学 Visible remote-sensing image sea-land segmentation method based on image segmentation and supervised learning
WO2015130231A8 (en) * 2014-02-27 2016-08-11 Agency For Science, Technology And Research Segmentation of cardiac magnetic resonance (cmr) images using a memory persistence approach

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102254303A (en) * 2011-06-13 2011-11-23 河海大学 Methods for segmenting and searching remote sensing image
CN102880872A (en) * 2012-08-28 2013-01-16 中国科学院东北地理与农业生态研究所 Classification and construction method for semi-supervised support vector machine (SVM) remote sensing image
WO2015130231A8 (en) * 2014-02-27 2016-08-11 Agency For Science, Technology And Research Segmentation of cardiac magnetic resonance (cmr) images using a memory persistence approach
CN104134219A (en) * 2014-08-12 2014-11-05 吉林大学 Color image segmentation algorithm based on histograms
CN105512622A (en) * 2015-12-01 2016-04-20 北京航空航天大学 Visible remote-sensing image sea-land segmentation method based on image segmentation and supervised learning

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
李勇发等: "基于FCM聚类及其改进的遥感图像分割算法", 《浙江农业科学》 *
来旭等: "基于半监督FCM聚类算法的卫星云图分类", 《国防科技大学学报》 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108830297A (en) * 2018-05-19 2018-11-16 烟台大学 A kind of multi-spectrum remote sensing image terrain classification method
CN108830297B (en) * 2018-05-19 2021-04-27 烟台大学 Multispectral remote sensing image ground feature classification method
CN111681245A (en) * 2020-06-17 2020-09-18 中原工学院 Method for segmenting remote sensing image based on tree structure adaptive weight k-means algorithm
CN111681245B (en) * 2020-06-17 2023-03-14 中原工学院 Method for segmenting remote sensing image based on tree structure adaptive weight k-means algorithm
CN111754501A (en) * 2020-06-30 2020-10-09 重庆师范大学 Self-adaptive soil image shadow detection method based on FCM algorithm
CN111754501B (en) * 2020-06-30 2021-08-27 重庆师范大学 Self-adaptive soil image shadow detection method based on FCM algorithm

Similar Documents

Publication Publication Date Title
CN107563381B (en) Multi-feature fusion target detection method based on full convolution network
CN106504255B (en) A kind of multi-Target Image joint dividing method based on multi-tag multi-instance learning
CN108985334A (en) The generic object detection system and method for Active Learning are improved based on self-supervisory process
CN103366367B (en) Based on the FCM gray-scale image segmentation method of pixel count cluster
CN102651128B (en) Image set partitioning method based on sampling
Zhang et al. PSO and K-means-based semantic segmentation toward agricultural products
CN112668579A (en) Weak supervision semantic segmentation method based on self-adaptive affinity and class distribution
CN102982544B (en) Many foreground object image interactive segmentation method
CN107730542A (en) Cone beam computed tomography image corresponds to and method for registering
CN106228554A (en) Fuzzy coarse central coal dust image partition methods based on many attribute reductions
CN101710418A (en) Interactive mode image partitioning method based on geodesic distance
CN105046714A (en) Unsupervised image segmentation method based on super pixels and target discovering mechanism
CN108846404A (en) A kind of image significance detection method and device based on the sequence of related constraint figure
CN105320764A (en) 3D model retrieval method and 3D model retrieval apparatus based on slow increment features
CN103678483A (en) Video semantic analysis method based on self-adaption probability hypergraph and semi-supervised learning
Yao et al. Sensing urban land-use patterns by integrating Google Tensorflow and scene-classification models
Chen et al. Agricultural remote sensing image cultivated land extraction technology based on deep learning
CN107452001A (en) A kind of remote sensing images sequences segmentation method based on improved FCM algorithm
CN109447111A (en) A kind of remote sensing supervised classification method based on subclass training sample
CN104573701B (en) A kind of automatic testing method of Tassel of Corn
CN102722578B (en) Unsupervised cluster characteristic selection method based on Laplace regularization
Bai et al. Calibrated focal loss for semantic labeling of high-resolution remote sensing images
CN113837191A (en) Cross-satellite remote sensing image semantic segmentation method based on bidirectional unsupervised domain adaptive fusion
CN106250828B (en) A kind of people counting method based on improved LBP operator
CN106250818B (en) A kind of total order keeps the face age estimation method of projection

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20171208

RJ01 Rejection of invention patent application after publication